Author(s)

Z. Piotrowski & P. Gajewski

Abstract

This paper refers to the technology of the acoustic watermark Internet server, which distributes digital sound together with additional inaudible hidden data. We describe the main architecture of this server (software modules) as well as the main features of this system. The characteristic specification referring to the implemented watermarking server’s engine is also presented. Data hiding is one of the most important technologies of Digital Rights Management dedicated for digital multimedia systems. Using hidden signals just below original host level and a perceptual model we can create another digital space for very important data. This data represents the index value for multimedia identification purposes (ID). In this paper we also describe experimental results using the internet infrastructure and dedicated algorithm for audio content fingerprinting. The implemented algorithm operates at the frequency domain and uses orthogonally distributed frequency components allocated along the spectrum bins. We also computed several metrics for a fixed number of audio tracks to show the watermark server and algorithm effectiveness. The following metrics were computed: the time of signal processing for watermark embedding and decoding, the Signal-to-Mask Ratio (SMR), the Mean Square Error (MSE), the Normalised Mean Square Error (NMSE), the Peak Signal to Noise Ratio (PSNR) and the Audio Fidelity (AF). Keywords: data hiding, watermarking, watermarked sound metrics, digital rights management.

Keywords

data hiding, watermarking, watermarked sound metrics, digital rights management.